0


有效的空间变异盲解卷积

Efficient space-variant blind deconvolution
课程网址: http://videolectures.net/nipsworkshops2010_harmeling_esv/  
主讲教师: Stefan Harmeling
开课单位: 马克斯普朗克研究所
开课时间: 2011-01-13
课程语种: 英语
中文简介:
由于相机抖动导致的照片模糊,大气湍流导致的天文图像序列模糊,以及物体运动导致的磁共振成像序列模糊,这些都是模糊的例子,不能充分描述为空间不变卷积,因为这种模糊在图像中会发生变化。在这篇文章中,我们提出了一类线性变换,它对于空间变量blurs有足够的表现力,但同时特别设计了有效的矩阵向量乘法。上述实例的成功结果说明了本文方法的实用意义。
课程简介: Blur in photos due to camera shake, blur in astronomical image sequences due to atmospheric turbulence, and blur in magnetic resonance imaging sequences due to object motion are examples of blur that can not be adequately described as a space-invariant convolution, because such blur varies across the image. In this talk, we present a class of linear transformations, that are expressive enough for space-variant blurs, but at the same time especially designed for efficient matrix-vector-multiplications. Successful results on the above-mentioned examples demonstrate the practical significance of our approach.
关 键 词: 卷积; 线性变换; 矩阵向量乘法
课程来源: 视频讲座网
最后编审: 2020-07-29:yumf
阅读次数: 48